JPH1188705A - Image reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To attain image reading with high quality by calculating an edge mount of a center pixel of a window set by a window setting means based on adjacent pixels, selecting a filter that calculates a plurality of output objects to select the output object thereby selectively enhancing the edge part. SOLUTION: A window setting means 14 sets part of areas where R, G, B output data from a sensor are written as a window independently of each color and the setting position is moved in the area. An edge amount calculation means 18 uses a center pixel of a matrix set by the window setting means 14 as an object pixel to calculate an edge amount of the object pixel in the window 16 and adjacent pixels. A filter selection means 19 selects a filter acting on the object pixel based on the calculated edge amount so as to operate an optimum filter. Thus, a side effect such as reproduction of moire is not caused and the extension of the skirt of the edge part is decreased.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image reading apparatus such as an image scanner or a digital copying machine for reading image information of a document.

[0002]

2. Description of the Related Art Conventionally, image reading apparatuses have been widely used as digital image input devices such as image scanners and copiers. In recent years, with the spread of the Internet and the need for high-quality digital images,
There is an increasing demand for colorization and higher image quality of the device.

FIG. 17 is a sectional view of a conventional image reading apparatus viewed from the front. In the figure, an image reading device 1
Are light source unit 4, reflection mirror group 5, imaging lens 6,
An optical carriage 8 configured as a reduction optical system by the line image sensor 7 and a drive motor 9 for driving the optical carriage 8 are provided.

A drive belt 12 is wound around a drive pulley 10 and a driven pulley 11 connected to a drive motor 9, and an optical carriage 8 is movably mounted on a guide shaft 13. Is moved, the original 2 on the original glass 3 is scanned to read the original image.

In such an image reading apparatus 1, when a user first sets a document 2 on a document glass 3 and a command to read the document is issued from an external host (not shown),
The light source unit 4 illuminates the document 2. The reflected light of the image information of the document 2 is reflected by the group of reflecting mirrors 5 fixed to the side wall of the optical carriage 8 and enters the imaging lens 6. Then, after the reflected light is scaled to a predetermined magnification by the imaging lens 6, an image is formed on a line image sensor 7 which is an element for converting optical information into an electric signal.

At the same time, the driving pulley 10, which has obtained the rotational force by the driving motor 9, transmits the driving force to the driving belt 12, which is given an appropriate tension by the driven pulley 11, and is fixed to the driving belt 12. The optical carriage 8 is guided by a guide shaft 13 fixed to a side wall of the image reading device 1 and scans a scanning range shown in FIG.

By such an operation, the image information on the document 2 is accumulated in the line image sensor 7 and all the image information of the document 2 can be taken. When the optical carriage 8 finishes scanning in the designated scanning range, it reverses its operation in the direction of arrow B and returns to the reading start position.
Stand by for reading the next original.

[0008]

In reading a color image, ideal luminance levels and scanning directions at the edges of R (red), G (green), and B (blue) sensors are as shown in FIG. It is well known that such a relationship is established. However, actually, due to the influence of the optical characteristics of the imaging lens 6, the vibration of the optical carriage 8, and the like, the spread W of the skirt field occurs at the edge portion of the read image as shown by the dashed arrow in FIG.

Due to the spread W of the skirt, there is a problem that the image near the edge becomes blurred and the quality of the image read by the scanner is reduced.

When the attribute of a region is determined, a low-pass filter is applied to the region determined to have the halftone attribute in order to prevent moire at the time of normal output. However, due to the low-pass filter, the width W of the skirt of the image edge portion becomes wider, and there is a problem that the quality of the image read by the scanner is further remarkably reduced.

For this reason, a method has been proposed in which a high-pass filter is applied once again to the entire image on which the low-pass filter is applied, thereby enhancing edges. However, in order to apply a uniform high-pass filter to the entire image, a low-pass filter is used. There is a problem that the moiré removed by the filter is regenerated.

An object of the present invention is to provide an image reading apparatus capable of reading a high-quality image by selectively changing a parameter depending on an edge amount and selectively enhancing an edge portion in image reading scanning. With the goal.

[0013]

According to the present invention, a part of an area where R, G, and B output data from a color image sensor is written is independently set as a window for each color, and the set position is set. A window setting means that can be moved within the area, and an edge amount of the target pixel from a target pixel in the window and a pixel adjacent to the target pixel using the center pixel of the window set by the window setting means as a target pixel. Edge amount calculation means for calculating; filter selection means for generating a signal for selecting a filter to be applied to the target pixel based on the edge amount calculated by the edge amount calculation means; and a target pixel and a target pixel in the window. Filter operation selecting means for calculating a plurality of output candidate values from data of adjacent pixels; From calculated by a plurality of output candidate value, characterized by comprising an output value selection means for selecting an output candidate value corresponding to the selection signal output from the filter selection means.

With such a configuration, the optimum filter can be selectively applied only to the edge portion of the read image, so that the spread of the hem of the edge portion is reduced without causing side effects such as re-generation of moire. Can be done.

[0015]

According to the first aspect of the present invention, a part of an area where each of R, G, and B output data from a color image sensor is written is set as a window independently for each color, and A window setting means capable of moving the setting position within the area; and a target pixel in the window and a pixel adjacent to the target pixel in the window, with the center pixel of the window set by the window setting means as the target pixel. Edge amount calculating means for calculating an edge amount; filter selecting means for generating a signal for selecting a filter to be applied to the target pixel based on the edge amount calculated by the edge amount calculating means; and a target pixel in the window. Filter operation selecting means for calculating a plurality of output candidate values from data of a pixel adjacent to the target pixel; Output value selecting means for selecting an output candidate value corresponding to a selection signal output from the filter selecting means from a plurality of output candidate values calculated in the stage, and selectively selecting an output part value suitable for an edge portion. Has the effect of reducing the spread of the skirt at the edge portion.

According to a second aspect of the present invention, a portion of an area where each of R, G, and B output data from the color image sensor is written is independently set as a window for each color, and the set position is set as the window. From the window setting means that can be moved within the area, and the target pixel in the window and the neighboring pixels of the target pixel, the edge amount of the target pixel is determined using the center pixel of the window set by the window setting means as the target pixel. An edge amount calculating means for calculating; a lookup table selecting means for generating a signal for selecting a lookup table for converting data of the target pixel based on the edge amount calculated by the edge amount calculating means; Lookup table reference that refers to multiple lookup table data from the target pixel data And output value selecting means for selecting an output candidate value corresponding to a selection signal output from the look-up table selecting means from a plurality of output candidate values referred to by the look-up table referring means. By selectively converting the data with the optimal look-up table for the edge part, the spread of the skirt of the edge part can be reduced, and the output data is not calculated, thereby improving the data throughput. It has the effect of doing.

According to a third aspect of the present invention, a part of an area in which R, G, and B output data from a color image sensor is written is set as a window independently for each color, and the set position is set in the area. Calculating the edge amount of the target pixel from the target pixel in the window and a pixel adjacent to the target pixel using the center pixel of the window set by the window setting unit as the target pixel An edge amount calculating unit that performs a threshold value selecting unit that generates a signal for selecting a threshold value to be attached to a maximum value or a minimum value of the luminance of the target pixel from the edge amount calculated by the edge amount calculating unit; Pasting means for pasting the target pixel to the maximum or minimum value of the luminance, and a plurality of output targets pasted by the pasting means. Output value selection means for selecting an output candidate value corresponding to the selection signal output from the threshold value selection means from among the values, and it is possible to reduce the spread of the skirt of the edge portion and to reduce the This has the effect that simplification can be achieved.

According to a fourth aspect of the present invention, a part of an area where each of the R, G, and B output data from the color image sensor is written is independently set as a window for each color, and the set position is set as the window. A window setting means that can be moved within the region, and an edge amount of the target pixel is calculated from a target pixel in the window and a pixel adjacent to the target pixel using the center pixel of the window set by the window setting means as a target pixel An edge amount calculating unit, an attribute determining unit that determines an attribute of the input image, and a target pixel based on the edge amount calculated by the edge amount calculating unit and a determination result of the attribute determining unit that determines an attribute of the input image. A filter selecting means for generating a signal for selecting a filter to be operated; and a target pixel in a window and a pixel adjacent to the target pixel. Filter operation selecting means for calculating the output candidate value of the above, and an output value for selecting an output candidate value corresponding to a selection signal output from the filter selecting means from a plurality of output candidate values calculated by the filter operation selecting means Selecting means for determining the halftone dot attribute by the attribute determining means and selectively applying an optimal high-pass filter to an edge portion of an area subjected to a low-pass filter to remove moiré; Thus, it is possible to reduce the spread of the skirt of the edge portion and to prevent the occurrence of moire.

The invention according to claim 5 provides the invention according to claims 1 to 4
The size of the window set by the window setting means is a 3 × 3 matrix, and the window can be configured with a minimum of hardware. This has the effect that simplification can be achieved.

The invention according to claim 6 is the invention according to claims 1 to 4
Wherein the directions of the edges calculated by the edge amount calculating means are four directions of the main scanning direction and the sub-scanning direction, and the edge amount calculating means may be configured with a minimum of hardware. Therefore, the circuit of the edge amount calculating means can be simplified.

According to a seventh aspect of the present invention, in the image reading apparatus of the fourth aspect, the attributes determined by the attribute determining means are three attributes of a character attribute, a halftone dot attribute, and a photograph attribute. Since the most suitable filter can be applied to each attribute based on the attribute determination result, it is necessary to perform processing to reduce the spread of the base of the character edge, the photograph attribute, and the halftone dot attribute corresponding to each attribute. It has the effect of being able to.

Hereinafter, embodiments of the present invention will be described. (Embodiment 1) FIG. 1 is a block diagram of a circuit for performing edge enhancement processing of an image reading apparatus according to Embodiment 1 of the present invention.

In FIG. 1, reference numeral 14 designates, as a window, a part of a region where an image, in which R, G, and B output data rm, gm, and bm from a color image sensor are written, is independently set for each color. Window setting means that can be moved within the area. The window setting means 14 includes an SRAM 15 for storing R, G, and B output data, a window 16 composed of three sets of 3 × 3 matrices for R, G, and B; It comprises a window RAM control means 17 for controlling the window 16. The window RAM control unit 17 receives a synchronization signal NHSZ in the main scanning direction, a synchronization signal NVSZ in the sub-scanning direction, and a pixel synchronization signal VCLK.

Reference numeral 18 denotes an edge amount calculating means for calculating the edge amount of the target pixel from the target pixel in the window 16 and a pixel adjacent to the target pixel, with the center pixel of the matrix set by the window setting means 14 as the target pixel. Yes, it is connected to the window setting means 14 by a bus.

Reference numeral 19 denotes a filter selecting unit for selecting a filter to be applied to the target pixel from the edge amount calculated by the edge amount calculating unit 18. The filter selecting unit 19 is connected to the edge amount calculating unit 18 by a bus.
Output Sel.

Reference numeral 20 denotes a filter calculation selecting means for calculating a plurality of output candidate values from data of pixels adjacent to the target pixel in the window 16, and this is a window setting means 1.
4 and a bus. Then, the filter operation selecting means 20 outputs the coordinates R00, R10, R20, G00, G10, G20,
B00, B10 and B20 are input data, and output candidate values fAR, fBR, fCR, fAG, fBG, fCG, f
AB, fBB, and fCB are output, respectively.

In the present embodiment, the filter operation selecting means 20 is composed of three types of filters, a filter A21, a filter B22, and a filter C23, and the second suffix of each output candidate value is the type of the filter. And the third subscript indicates the output color channel.

Further, reference numeral 24 denotes a filter selection means 1 based on a plurality of output candidate values calculated by the filter calculation selection means 20.
Reference numeral 9 denotes output value selection means for selecting an output candidate value corresponding to the filter selection signal Filter-Sel output.

FIG. 3 is a diagram showing a configuration of the edge amount calculating means 18 according to the first embodiment of the present invention, FIG. 4 is a diagram showing an edge amount calculating operator in the upward, downward, left and right directions, and FIG. Is a diagram showing a configuration of the filter selection means 19,
FIG. 6 is a diagram showing the coordinates of the same image, FIG. 7 is a diagram showing an example of the configuration of the filter operation selection means 20, and FIG. 8 is a diagram showing the configuration of the output value selection means 24.

The operation of the image reading apparatus having the above configuration will be described below with reference to these drawings. Outputs from the line image sensor 7 include a synchronization signal NHSZ in the main scanning direction, a synchronization signal NVSZ in the sub-scanning direction, and a pixel synchronization signal VCLK.
Are transmitted to the window setting means 14 in synchronism with R00, R10, R2 of the window 16 shown in FIG.
0, G00, G10, G20, B00, B10, B20
Is stored in At the same time, the image data of one line and two lines before that is held in the SRAM 15 is read to form a window. At this time, R00, R10, R2
0, G00, G10, G20, B00, B10, B20
Is necessary for forming a window with the next line as a target pixel, so writing to the SRAM 15 is also performed.

The window 16 thus constructed
On the other hand, the edge amount calculating means 18 has the configuration shown in FIG. 3, and allows the upward, downward, leftward, and rightward edge amount calculating operators shown in FIG. Thus, the edge amount in each plane and each direction is obtained.

In the actual calculation of the edge detection operator in FIG. 3, the product of the value of the operator at the same position and the value of the window 16 is calculated, and the calculation is performed to sum them. This will be specifically described below using the R window 16 (R) as an example.

First, the upward edge amount is UlobR, the downward edge amount is DrobR, and the leftward edge amount is Lr.
obR, the edge amount in the right direction is RrobR. Then, the edge amount calculation means 18 performs the following calculation to obtain the edge amount in each direction.

UrobR = R00 + 2R01 + R02- (R20 + 2R21 + R22)... (1) DrobR =-(R00 + 2R01 + R02) + R20 + 2R21 + R22. ) + R02 + 2R12 + R22 (4) Similarly, from the value of the G window 16 (G), UrobG,
From the values of the window 16 (B) of B, UrobB, DrobB, L
An operation for obtaining robB and RrobB is performed.

Next, the filter selecting means 19 is constructed as shown in FIG.
R, LlobR, RlobR, UrobG, Drob
G, LlobG, RlobG, UrobB, Drob
B, LlobB, RlobB, and edge determination threshold signals Edg-th1 to Edg-th3 are input signals to output a filter selection signal Filter-Sel. The filter selection signal Filter-Sel is a selection signal Fi that does not select an output candidate value for each of the R, G, and B channels.
lter-SelR0, Filter-SelG0, F
selection signals Filter-SelR1 and Filter-Se for selecting filter-SelB0 and filter A21.
Selection signals Filter-SelR2, Filter for selecting 1G1, Filter-SelB1, and Filter B22
er-SelG2, Filter-SelB2, and a selection signal Filter-Sel for selecting the filter C23.
R3, Filter-SelG3, Filter-Se
1B3.

The filter selecting method of the filter selecting means 19 is as follows. First, MaxrobR, MaxrR are set as the maximum edge amounts of the R, G, and B channels of the target pixel.
ObG and MaxrobB are calculated respectively.

Next, for the R channel, Max
If the value of robR is less than the edge determination threshold value Edg-th1, that is, if Edg-th1> MaxrobR, a selection signal Filter-S that does not select an output candidate value
elR0 is asserted.

When the value of MaxrobR is equal to or more than the edge determination threshold value Edg-th1 and less than Edg-th12,
That is, Edg-th1 ≦ MaxrobR <Edg-t
h2, the selection signal F for selecting the filter A21
ilter-SelR1 is asserted. Also, Ma
When the value of xrobR is equal to or more than the edge determination threshold value Edg-th2 and less than Edg-th3, that is, Edg-th2 ≦
When MaxrobR <Edg-th3, the selection signal Filter-SelR2 for selecting the filter B22 is asserted. Further, when the value of MaxrobR is equal to or greater than the edge determination threshold value Edg-th3, that is, when Edg-t
When h3 ≦ MaxrobR, the selection signal Filter-SelR3 for selecting the filter C23 is asserted.

Similarly, for the G channel, Max
According to the value of robG, Filter-SelG0, Fi
lter-SelG1, Filter-SelG2, F
Any one of the filter-SelG3 also determines the Filter-
SelB0, Filter-SelB1, Filter
One of -SelB2 and Filter-SelB3 is asserted.

The filter operation means 20 has the configuration shown in FIG. 6, and the filter operation means in the illustrated example performs the same operation as that performed by the edge amount calculation means. Specifically, the calculation is performed by multiplying the value of the operator at the same position of each filter shown in FIG. 7 by the value of the window 16 and taking the sum thereof. Specifically, an R window 16 (R) will be described as an example.

Assuming that the output candidate value of the filter A of the R channel is fRA, that of the filter B is fBR, and that of the filter C is fCR, the filter calculation output means 20 calculates the following output candidate value of each filter. Perform the operation.

FAR = 2R11- ／ (R01 + R10 + R12 + R21) (5) fBR = 3R11-1 / 2 (R01 + R10 + R12 + R21) (6) fCR = R11- (R01 + R10 + R12 + R21) (7) Similarly FAG, fB from the value of window 16 (G) of G
G and fCG are calculated from the value of the window 16 (B) of B by fA
An operation for obtaining B, fBB, and fCB is performed.

Finally, the output value selecting means 24 has the configuration shown in FIG.
R, fBR, fCR, fAG, fBG, fCG, fA
Selection of output candidate values corresponding to the filter selection signal output by the filter selection means 19 for each of the R, G, B channels from the output data rm, gm, bm of the color image sensor 7 for B, fBB, fCB. Do
CCDA-R, CCDA-G, CCD as output data
Output AB.

A specific description will be given by taking the R channel as an example.
When the filter selection signal Filter-SelR0 is asserted, rm is output as CCDA-R. Further, when Filter-SelR1 is asserted, fAR, and when Filter-SelR2 is asserted, fBR is Filter-SelR2.
If SelR3 is asserted, fBC
Output as CDA-R.

For the G channel and the B channel, in the same manner, the output candidate values corresponding to the asserted signal in the filter selection signal are respectively set to CCDA.
-G, output as CCDA-B.

Second Embodiment Next, a second embodiment of the present invention will be described.

FIG. 9 is a block diagram of a circuit for performing edge enhancement processing of the image reading apparatus according to the second embodiment. The same components as those in the image reading apparatus according to the first embodiment of the present invention shown in FIG. And a description thereof will be omitted.

This image reading apparatus differs from the first embodiment in FIG. 1 in that the filter selecting means 19 is replaced by a look-up table selecting means 25 and the filter calculating means 20 is replaced by a look-up table referring means 26. Is a point.

In this embodiment, the look-up table reference means 26 is composed of three types of filters, a table A27, a table B28 and a table C23.
The second suffix of each table reference value indicates the type of the table, and the third suffix indicates the output color channel.

FIG. 10 is a diagram showing a configuration of the lookup table selecting means 25 according to the second embodiment of the present invention.
1 is a diagram showing a configuration of the lookup table reference means 26.

The look-up table selecting means 25 has the configuration shown in FIG. 10 and includes UlobR, DrobR, Lr
obR, RobR, UrobG, DrobG, Lro
bG, RlobG, UrobB, DrobB, Llob
B, RlobB, edge determination threshold signal Edg-th1
A lookup table selection signal Table-Sel is output using Edg-th3 as an input signal. The filter selection signals Table-Sel are selection signals Table-SelR0, Table-Sel that do not select the lookup table reference value for each of the R, G, and B channels.
Selection signals Table-SelR1, Table-S for selecting G0, Table-SelB0, and Table A27
elG1, Table-SelB1, selection signal Table-SelR2, Table for selecting table B28
-SelG2, Table-SelB2, Table C2
9 selecting signal Table-SelR3, Tab
It is composed of le-SelG3 and Table-SelB3.

The filter selection method of the lookup table selection means 25 is the same as that of the filter selection means described in the first embodiment, and the maximum edge amounts MaxrobR, MaxrR of the R, G, and B channels of the target pixel are used.
ObG and MaxrobB are obtained.

Next, for the R channel, Maxr
When the value of obR is less than the edge determination threshold value Edg-th1, that is, when Edg-th1> MaxrobR, the selection signal Table-SelR0 that does not select the reference value of the lookup table is asserted. Also, M
The value of axrobR is E at the edge determination threshold value Edg-th1.
dg-th2, that is, Edg-th1 ≦ M
When axrobR <Edg-th2, the selection signal Table-Se for selecting the lookup table A27 is used.
lR1 is asserted. Further, when the value of MaxrobR is equal to or more than the edge determination threshold value Edg-th2 and less than Edg-th3, that is, Edg-th2 ≦ MaxrobR <
In the case of Edg-th3, the selection signal Table-SelR2 for selecting the table B28 is asserted. Further, the value of MaxrobR is the edge determination threshold value Edg-th
In the case of 3 or more, that is, Edg-th3 ≦ Maxrob
In the case of R, a selection signal Ta for selecting the table C29
ble-SelR3 is asserted.

Similarly, for the G channel, Max
Depending on the value of robG, Table-SelG0, Tab
le-SelG1, Table-SelG2, Tabl
Any of the e-SelG3s can also determine the value of the Max-RobB for the channel of the Table-SelG3.
0, Table-SelB1, Table-SelB
2, either Table-SelB3 is asserted.

Also, look-up table reference means 26
Has a configuration shown in FIG. 11, and R11, G11,
By referring to each look-up table with B11 as an input value, tAR, tBR, tCR, tAG, t
BG, tCG, tAB, tBB, and tCB are output, respectively.

Finally, the output value selection means 24 outputs the table reference values tAR, tBR, tCR, tAG, tB of the lookup table reference means 26, as in the first embodiment.
From G, tCG, tAB, tBB, tCB, and output data rm, gm, bm of the color image sensor 7, R,
For each of the channels G and B, an output candidate value corresponding to the lookup table selection signal output by the lookup table selection means 25 is selected, and CCDA-R, CCDA-G, and CCDA-B are output as output data. I do.

(Embodiment 3) FIG. 12 is a block diagram of a circuit for performing edge enhancement processing of an image reading apparatus according to Embodiment 3 of the present invention. For the same components as the image reading device,
The same reference numerals are given, and the description is omitted.

This image reading apparatus is different from the first embodiment in that the filter selecting means 19 is attached to the threshold value selecting means 30 and the filter calculating means 20 is attached to the attaching means 31 for attaching to the maximum or minimum luminance value. It has been replaced.

In the present embodiment, the threshold value selecting means 30
It consists of three types of thresholds, threshold A, threshold B, and threshold C. The second suffix of the output candidate value of the pasting means indicates the type of the table, and the third suffix indicates the color channel of the output. And

FIG. 13 is a diagram showing the structure of the threshold value selecting means 30 according to the third embodiment of the present invention, and FIG.

The threshold value selecting means 30 has the configuration shown in FIG. 13 and includes UrobR, DrobR, LlobR, and Rlob.
R, UrobG, DrobG, LlobG, Rlob
G, UrobB, DrobB, LlobB, Rlob
B, edge determination threshold signals Edg-th1 to Edg-th
3 as an input signal, and outputs a threshold selection signal Th-Sel. The threshold selection signal Th-Sel is a selection signal Th-SelR0, Th-SelG0, Th that does not select an output candidate value of the pasting means for each of R, G, and B channels.
-SelB0, selection signal Th-Sel for selecting threshold A
R1, Table-SelG1, Table-SelB
1. Selection signal Th-SelR2, Th for selecting threshold B
Selection signals Th-SelR3, Th-SelG3, Th-S for selecting SelG2, Th-SelB2, and threshold C
elB3.

The threshold selection method of the threshold selection unit 30 is the same as that of the filter selection unit of the first embodiment, and the maximum edge amount M of each of the R, G, and B channels of the target pixel is set.
Find axrobR, MaxrobG, and MaxrobB.

Next, for the R channel, Maxr
When the value of obR is smaller than the edge determination threshold value Edg-th1, that is, when Edg-th1> MaxrobR, the selection signal T that does not select the output candidate value of the pasting unit.
h-SelR0 is asserted. Also, Maxrob
When the value of R is equal to or greater than the edge determination threshold value Edg-th1, Edg-
less than th2, ie, Edg-th1 ≦ Maxr
When obR <Edg-th2, the selection signal Table-SelR1 for selecting the threshold A is asserted. Further, the value of MaxrobR is the edge determination threshold value Edg-th
2 or more and less than Edg-th3, that is, Edg-th
When th2 ≦ MaxrobR <Edg-th3,
The selection signal Th-SelR2 for selecting the threshold B is asserted. Also, the value of MaxrobR is equal to the edge determination threshold E
dg-th3 or more, that is, Edg-th3 ≦ M
In the case of axrobR, the selection signal T for selecting the threshold value C
h-SelR3 is asserted.

Similarly, for the G channel, Max
According to the value of robG, Th-SelG0, Th-Sel
Any of G1, Th-SelG2, and Th-SelG3 also determines Th-SelB0, Th-SelB1, and Th-Se for the B channel according to the value of MaxlobB.
One of IB2 and Th-SelB3 is asserted.

The pasting means 31 has the configuration shown in FIG. 14, and uses R11, G11, and B11 as input values.
Pasting to the maximum value or minimum value of the luminance is performed according to the threshold, and hAR, hBR, hCR, hAG, hBG, hC
G, hAB, hBB, and hCB are output.

The method of pasting by the pasting means will be specifically described by taking the R channel as an example.
Is less than the threshold value A, that is, when R11 <threshold value A, h
AR is pasted on the minimum luminance value Ymin. vice versa,
When the threshold value A or more, that is, when R11 ≧ threshold value A, hA
R is pasted on the maximum luminance value Ymax.

Similarly, the same paste is performed for the threshold B, the threshold C, and the channels G and B.

Finally, the output value selecting means 24 outputs the output candidate values hAR, h of the pasting means 31 in the same manner as in the first embodiment.
BR, hCR, hAG, hBG, hCG, hAB, hB
From the B, hCB, and output data rm, gm, bm of the color image sensor 7, selection of an output candidate value corresponding to the threshold selection signal output by the output threshold selection unit 25 for each of the R, G, B channels. Then, CCDA-R, CCDA-G, and CCDA-B are output as output data.

(Embodiment 4) FIG. 15 is a block diagram of a circuit for performing edge enhancement processing of an image reading apparatus according to Embodiment 4 of the present invention. The image according to Embodiment 1 of the present invention shown in FIG. The same components as those of the reading device are denoted by the same reference numerals, and description thereof will be omitted.

This image reading apparatus is different from the first embodiment in FIG. 1 in that an attribute judging means 35 for judging the attribute of an input image is added, and the attribute judgment result is inputted to a filter selecting means 36. , And the filter operation means 37 is provided with a filter D38 and a filter E39 in addition to the filters A21, B22 and C23, and the output candidate values fAR, fBR, fCR, fAG,
fBG, fCG, fAB, fBB, fCB, fDR, f
DG, fDB, fER, fEG, and fEB are output. In the present embodiment, the number of filters is five, but any number of filters may be used as long as the number of attributes is equal to or larger than the number of attributes for attribute determination.

When the attribute determination result is a character determination,
No filter, one of filter A21, filter B22, and filter C23 are selected. In the case of photo determination, no filter, filter B22, filter C2
3 and the filter D38 are selected, and in the case of the halftone dot determination, there is no filter, the filter C23, the filter D
38 and the filter E39 are selected.

The attribute determining means 35 determines the attribute of the input image and outputs an attribute determination signal. The attribute determination signal includes a character determination Char-Sig, a halftone determination Ami-Sig, and a photo determination Photo-Sig. If the determination result is a character determination, Char-Sig is determined. If the determination result is a halftone dot, Ami-S is determined.
If ig is a photo determination, Photo-Sig is asserted.

The filter selecting means 36 is constructed as shown in FIG. 16 and includes UlobR, DrobR, LlobR, R
robR, UrobG, DrobG, LlobG, Rr
obG, UrobB, DrobB, LlobB, Rro
bB, edge determination threshold signals Edg-th1 to Edg-t
h3, attribute determination signals Char-Sig, Ami-Si
g, Photo-Sig as an input signal, and outputs a filter selection signal Filter-Sel. The filter selection signal Filter-Sel is a selection signal Filter that does not select an output candidate value for each of the R, G, and B channels.
r-SelR0, Filter-SelG0, Filt
er-SelB0, selection signals Filter-SelR1, Filter-SelG for selecting the filter A21.
1, a selection signal Filter-SelR2 for selecting Filter-SelB1 and a filter B22, Filter
-SelG2, Filter-SelB2, Filter C
Selection signal Filter-SelR3, F for selecting
ilter-SelG3, Filter-SelB3,
Selection signal Filter-Se for selecting filter D38
IR4, Filter-SelG4, Filter-S
elB4, selection signal Filt for selecting filter E39
er-SelR5, Filter-SelG5, Fil
ter-SelB5.

The filter selection method of the filter selection means 36 is the same as that of the first embodiment.
Maximum edge amount Maxr of each channel of G and B
ObR, MaxrobG, and MaxrobB are obtained.

Next, for the R channel, if the attribute determination result is a character determination, that is, Char-Sig is asserted, and the value of MaxrobR is less than the edge determination threshold value Edg-th1, that is, Edg-th1> Max
In the case of robR, a selection signal Filter-SelR0 that does not select an output candidate value is asserted. Also, M
The value of axrobR is equal to or more than the edge determination threshold value Edg-th1 and less than Edg-th12, that is, Edg-th1 ≦ M
If axrobR <Edg-th2, filter A
A selection signal Filter-SelR1 for selecting 21 is asserted.

When the value of MaxrobR is equal to or larger than the edge determination threshold value Edg-th2 and smaller than Edg-th3, that is, when Edg-th2 ≦ MaxrobR <Edg-th3, the selection signal Filter for selecting the filter B22 is selected.
r-SelR2 is asserted. Also, Maxrob
When the value of R is equal to or greater than the edge determination threshold value Edg-th3, that is, when Edg-th3 ≦ MaxrobR, the selection signal Filter-SelR3 for selecting the filter C23 is asserted.

If the attribute determination result is a photo determination, that is, Photo-Sig is asserted and Maxrob
When the value of R is less than the edge determination threshold value Edg-th1, that is, when Edg-th1> MaxrobR, a selection signal Filter-SelR0 that does not select an output candidate value is asserted. When the value of MaxrobR is equal to or more than the edge determination threshold value Edg-th1 and less than Edg-th2, that is, when Edg-th1 ≦ MaxrobR <Edg-th2, the selection signal Filt for selecting the filter B22 is selected.
er-SelR2 is asserted. Also, Maxro
When the value of bR is equal to or greater than the edge determination threshold value Edg-th2, Edg
Less than −th3, that is, Edg−th2 ≦ Maxrob
When R <Edg-th3, the selection signal Filter-SelR3 for selecting the filter C23 is asserted. Also, the value of MaxrobR is the edge determination threshold value Edg
−th3 or more, ie, Edg-th3 ≦ Max
In the case of robR, a selection signal Filter-SelR4 for selecting the filter D38 is asserted.

When the attribute determination result is a halftone dot determination, that is, Ami-Sig is asserted and Maxro
When the value of bR is less than the edge determination threshold value Edg-th1, that is, when Edg-th1> MaxrobR, a selection signal Filter-SelR0 that does not select an output candidate value is asserted. When the value of MaxrobR is equal to or more than the edge determination threshold value Edg-th1 and less than Edg-th2, that is, when Edg-th1 ≦ MaxrobR <Edg-th2, the selection signal Filt for selecting the filter C23 is selected.
er-SelR3 is asserted. Also, Maxro
When the value of bR is equal to or greater than the edge determination threshold value Edg-th2, Edg
Less than −th3, that is, Edg−th2 ≦ Maxrob
When R <Edg-th3, the selection signal Filter-SelR4 for selecting the filter D38 is asserted. Also, the value of MaxrobR is the edge determination threshold value Edg
−th3 or more, ie, Edg-th3 ≦ Max
In the case of robR, a selection signal Filter-SelR5 for selecting the filter E39 is asserted.

Similarly, for the G and B channels, M
Filter-SelG0, Filter-SelG0, Filro-SelG0, Filro-SelG0, Filro-SelG0
ter-SelG1, Filter-SelG2, Fi
lter-SelG3, Filter-SelG4, F
ilter-SelG5, Filter-SelB0,
Filter-SelB1, Filter-SelB
2, Filter-SelB3, Filter-Sel
One of B4 and Filter-SelB5 is asserted.

Finally, in the output value selecting means 24, the output candidate values fAR, fBR, fCR,
fDR, fER, fAG, fBG, fCG, fDG, f
EG, fAB, fBB, fCB, fDB, fEB, and
Output data rm, gm, bm of color image sensor 7
, For each of the R, G, and B channels, an output candidate value corresponding to the filter selection signal output by the filter selection means 36 is selected, and CCDA is output as output data.
-R, CCDA-G and CCDA-B are output.

More specifically, taking the R channel as an example,
When the filter selection signal Filter-SelR0 is asserted, rm is output as CCDA-R. Further, when Filter-SelR1 is asserted, fAR, and when Filter-SelR2 is asserted, fBR is Filter-SelR2.
When SelR3 is asserted, fBC becomes FBC
When filter-SelR4 is asserted, fDR is output as CCDA-R. When Filter-SelR5 is asserted, fER is output as CCDA-R.

For the G channel and the B channel,
In the same process, the output candidate values corresponding to the asserted signal in the filter selection signal are respectively output from the CCDA-
G, output as CCDA-B.

[0083]

According to the first aspect of the present invention, an optimum high-pass filter can be selectively applied to an edge portion, so that it is possible to reduce the spread of the hem field of the edge portion, thereby obtaining a read image. Image quality can be improved.

According to the second aspect of the present invention, the data is selectively converted to the edge portion by using the optimum look-up table, so that the spread of the hem field at the edge portion can be reduced and the image quality can be improved. Since no calculation is performed, data throughput can be improved, and it is possible to easily cope with high-speed equipment.

According to the third aspect of the present invention, instead of using relatively complicated hardware such as filter operation means and look-up table reference means, the maximum value and minimum value of the luminance are simple. Since the device can be constituted by the attaching means, the spread of the skirt of the edge portion can be reduced, the circuit can be simplified, and the cost can be reduced.

According to the fourth aspect of the present invention, the attribute determining means determines the halftone dot attribute and applies an optimal high-pass filter selectively to an edge portion of a region subjected to the low-pass filter to remove moire. By doing so, it is possible to improve the image quality by reducing the spread of the skirt of the edge portion, and it is possible to prevent the re-generation of moire.

According to the fifth aspect of the present invention, since the window is constituted by the minimum hardware, the circuit of the window setting means can be simplified, so that an increase in the apparatus cost can be minimized.

According to the sixth aspect of the present invention, since the edge amount calculation means is constituted by the minimum hardware, the edge amount calculation means can be simplified, so that an increase in apparatus cost can be minimized. it can.

According to the seventh aspect of the present invention, different processing corresponding to each attribute can be performed according to the attribute determination result, so that the quality of a read image is further improved.

[Brief description of the drawings]

FIG. 1 is a block diagram of a circuit that performs edge enhancement processing of an image reading device according to a first embodiment of the present invention.

FIG. 2 is a diagram showing coordinates of a window according to the first embodiment of the present invention.

FIG. 3 is a diagram showing a configuration of an edge amount calculating unit according to the first embodiment of the present invention;

FIG. 4 is a diagram showing an edge amount calculation operator in an upward direction, a downward direction, a left direction, and a right direction according to the first embodiment of the present invention;

FIG. 5 is a diagram showing a configuration of an edge determination unit according to the first embodiment of the present invention.

FIG. 6 is a diagram showing coordinates of an image according to the first embodiment of the present invention;

FIG. 7 is a diagram illustrating an example of a configuration of a filter operation unit according to the first embodiment of the present invention.

FIG. 8 is a diagram showing a configuration of an output value selection unit according to the first embodiment of the present invention.

FIG. 9 is a block diagram of a circuit that performs edge enhancement processing of the image reading device according to the second embodiment of the present invention;

FIG. 10 is a diagram showing a configuration of a look-up table selecting unit according to the second embodiment of the present invention.

FIG. 11 is a diagram showing a configuration of a lookup table reference unit according to the second embodiment of the present invention.

FIG. 12 is a block diagram of a circuit that performs edge enhancement processing of the image reading device according to the third embodiment of the present invention.

FIG. 13 is a diagram showing a configuration of a threshold value selecting unit according to the third embodiment of the present invention.

FIG. 14 is a diagram showing a configuration of a pasting unit according to Embodiment 3 of the present invention.

FIG. 15 is a block diagram of a circuit that performs edge enhancement processing of the image reading device according to the fourth embodiment of the present invention.

FIG. 16 is a diagram showing a configuration of a filter selection unit according to a fourth embodiment of the present invention.

FIG. 17 is a cross-sectional view of a conventional image reading apparatus as viewed from the front.

FIG. 18 is a diagram illustrating a relationship between a luminance level of an edge portion of each of ideal R, G, and B sensors and a scanning direction;

FIG. 19 is a diagram showing the relationship between the luminance level of the edge portion of each of the R, G, and B sensors and the scanning direction.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Image reading apparatus 2 Document 3 Document glass 4 Light source unit 5 Reflection mirror group 6 Imaging lens 7 Line image sensor 8 Optical carriage 9 Drive motor 10 Drive pulley 11 Follower pulley 12 Drive belt 13 Guide shaft 14 Window setting means 15 SRAM 16 Window 17 Window RAM control means 18 Edge amount calculation means 19 Filter selection means 20 Filter calculation selection means 21 Filter A 22 Filter B 23 Filter C 24 Output value selection means

Claims (7)

[Claims] A part of an area where R, G, and B output data from a color image sensor is written is set as a window independently for each color, and the set position is moved within the area. Possible window setting means, and edge amount calculation means for calculating an edge amount of a target pixel from a target pixel in the window and a pixel adjacent to the target pixel with the center pixel of the window set by the window setting means as a target pixel, Filter selecting means for generating a signal for selecting a filter to be applied to the target pixel based on the edge amount calculated by the edge amount calculating means; and a plurality of data from the data of the target pixel and the adjacent pixel to the target pixel in the window. Filter operation selecting means for calculating an output candidate value, and a plurality of output candidate values calculated by the filter operation selecting means Al, the image reading apparatus characterized by comprising an output value selection means for selecting an output candidate value corresponding to the selection signal output from the filter selection means. 2. A method according to claim 1, wherein a part of an area in which R, G, B output data from the color image sensor is written is set as a window independently for each color, and the set position is moved in said area. Possible window setting means, and edge amount calculating means for calculating an edge amount of the target pixel from a target pixel in the window and a pixel adjacent to the target pixel with the center pixel of the window set by the window setting means as a target pixel. A look-up table selecting means for generating a signal for selecting a look-up table for converting the data of the target pixel based on the edge amount calculated by the edge amount calculating means; Lookup table reference means for referring to the data of the lookup table;
Output value selecting means for selecting an output candidate value corresponding to a selection signal output from the look-up table selecting means from a plurality of output candidate values referred to by the look-up table referring means. Image reading device. 3. A part of an area where R, G, and B output data from a color image sensor is written can be set as a window independently for each color, and the set position can be moved within the area. A window setting means, and an edge amount calculating means for calculating an edge amount of the target pixel from a target pixel in the window and a pixel adjacent to the target pixel using the center pixel of the window set by the window setting means as a target pixel; Threshold value selecting means for generating a signal for selecting a threshold value to be attached to the maximum value or the minimum value of the luminance of the target pixel from the edge amount calculated by the edge amount calculating means; Or pasting means for pasting to the minimum value, and threshold selecting means from among a plurality of output candidate values pasted by the pasting means Image reading apparatus characterized by comprising an output value selection means for selecting an output candidate value corresponding to the selection signal output Ri. 4. A part of an area in which output data of each of R, G, and B from a color image sensor is written is set as a window independently for each color, and the set position is moved within the area. Possible window setting means, and edge amount calculation means for calculating an edge amount of a target pixel from a target pixel in the window and a pixel adjacent to the target pixel with the center pixel of the window set by the window setting means as a target pixel, A signal for selecting a filter to be applied to the target pixel based on the edge amount calculated by the edge amount calculating means and the determination result of the attribute determining means for determining the attribute of the input image; And a plurality of output candidate values are calculated from the target pixel in the window and the neighboring pixels of the target pixel. Filter operation selecting means, and output value selecting means for selecting an output candidate value corresponding to a selection signal output from the filter selecting means from a plurality of output candidate values calculated by the filter operation selecting means. An image reading device characterized by the above-mentioned. 5. The image reading apparatus according to claim 1, wherein the size of the window set by the window setting means is a 3 × 3 matrix. 6. The image reading apparatus according to claim 1, wherein the direction of the edge calculated by the edge amount calculating means is four directions of a main scanning direction and a sub-scanning direction. 7. The image reading apparatus according to claim 4, wherein the attributes determined by the attribute determining means are three attributes of a character attribute, a halftone dot attribute, and a photograph attribute.